Extracting Dynamical Correlations and Identifying Key Residues for Allosteric Communication in Proteins by correlationplus

Mustafa Tekpinar (1) , Bertrand Neron (2) , Marc Delarue (1)

1 Architecture et Dynamique des Macromolécules Biologiques - Architecture and Dynamics of Biological Macromolecules
2 Hub Bioinformatique et Biostatistique - Bioinformatics and Biostatistics HUB

Extracting dynamical pairwise correlations and identifying key residues from large molecular dynamics trajectories or normal-mode analysis of coarse-grained models are important for explaining various processes like ligand binding, mutational effects, and long-distance interactions. Efficient and flexible tools to perform this task can provide new insights about residues involved in allosteric regulation and protein function. In addition, combining and comparing dynamical coupling information with sequence coevolution data can help to understand better protein function. To this aim, we developed a Python package called correlationplus to calculate, visualize, and analyze pairwise correlations. In this way, the package aids to identify key residues and interactions in proteins. The source code of correlationplus is available under LGPL version 3 at https://github.com/tekpinar/correlationplus. The current version of the package (0.2.0) can be installed with common installation methods like conda or pip in addition to source code installation. Moreover, docker images are also available for usage of the code without installation.

Mots clés: Allostery; coevolution; coupling; dynamical cross-correlation; elastic network model; linear mutual information; molecular dynamics simulations; Allostery

Domaines: Cristallographie; Bio-informatique [q-bio.QM]; Biophysique [physics.bio-ph]; Biochimie, Biologie Moléculaire; Biologie structurale [q-bio.BM]; Biologie cellulaire

Mustafa Tekpinar, Bertrand Neron, Marc Delarue. Extracting Dynamical Correlations and Identifying Key Residues for Allosteric Communication in Proteins by correlationplus. Journal of Chemical Information and Modeling, 2021, 61 (10), pp.4832-4838. ⟨10.1021/acs.jcim.1c00742⟩. ⟨pasteur-03864465⟩